Demo abstract: Toward optimal allocation of multiple power resources in Energy-on-Demand systems

Abstract

Future power networks are supposed to contain various power sources in mesh-like structures, including distributed generations such as solar panels, wind power generations, and electric vehicles (EVs) that work not only as transportation but as huge batteries. These power sources have their own characteristics such as cost, stability, amounts of CO2 emission, etc. Hence it is desired to use them efficiently in a way suited for various characteristics of power consuming devices; for example, a desktop computer should be supplied with power from sufficiently stable power sources such as a commercial power source, while a laptop computer with a battery may accepts less stable power from natural power sources. In the concept of Energy-on-Demand (EoD), Quality-of-Energy (QoEn) power routing has been proposed to realize end-to-end power routing between power sources and power consuming device for efficient use of various power sources to optimize power allocation. This abstract describes the design and prototype implementation of the EoD system with multiple power resources utilizing power allocation management based on the Multiple Knapsack Problem with Assignment Restrictions (MK-AR), a kind of the combinatorial optimization problem.